142 
MANGANESE DEPOSITS OP INDIA : MINERALOGY. [ PaRT I : 
is never found in the typical members of the kodurite series of Vizaga- 
patam, but only in those varieties I have designated the manganese- 
pyroxenites (see page 249). These, however, are nearly always found 
apart from the typical kodurite rocks, so that at first sight it might be 
thought that the mere fact of their containing rhodonite, considering 
what has been written above, might be taken to indicate that these 
manganese-pyroxenites are also of metamorphic origin and form a 
distinct series from the igneous kodurite series. It seems more 
probable, however, that the manganese-pyroxenites are also of igneous 
origin, since they are sometimes found in association with the typical 
kodurite rocks, as at Perapi and Kodur, for instance ; and that 
the absence of rhodonite in the typical kodurite rocks is due to the fact 
that rhodonite requires about 42 per cent, of manganese for the 
formation of pure MnSiOs ; and that when many other constituents 
are present in the rock, so that the percentage of manganese is 
relat vely low, the manganese naturally forms silicates that contain 
a smaller percentage of manganese, such as the other manganiferous 
pyroxenes or manganese-garnets. The abundance of rhodonite in the 
manganiferous rocks of metamorphic origin is then explained by the 
frequency with which the original sediments from which the rocks of 
the gondite series were formed contained a high percentage of manganese. 
The same reasoning as for Vizagapatam explains why rhodonite is not 
formed in the rocks intrusive into the gondite series. 
As already mentioned, rhodonite exhibiting measurable crystal 
faces has never been fc.und in the Indian deposits, although such crystals 
are sometimes found in other parts of the world. This is probably due 
to its usual metamorphic origin, the rocks in which it occurs having been 
subjected to great pressures at the time of formation, so that as a rule 
they contain no open spaces or cavities. Such spaces as there are, 
have been formed by the action of waters subsequent to the 
formation of the rhodonite, so that if any such empty spaces are lined 
with crystals, they are of subsequently formed minerals, such as 
the various oxide minerals, rather than of rhodonite. This can- 
not, however, be the whole of the explanation why rhodonite crystals 
are not found, because the manganese-garnet, spessartite, associated 
with the rhodonite and formed under the same circumstances, frequently 
occurs in well-formed crystals, usually in a matrix of quartz, but 
sometimes in a matrix of rhodonite. Hence it seems as if rhodonite has 
very little tendency to take on well-crystallized forms. In the rocks 
of the gondite series it sometimes occurs as rhodonite-rock, forming 
